Information About Graceful Insertion and Removal
Overview
Graceful Insertion and Removal (GIR) isolates a switch from the network in order to perform debugging or an upgrade. The switch can be put into maintenance mode using the start maintenance command. When switch maintenance is complete, the switch will return to normal mode on either reaching the configured maintenance timeout, or by enabling the stop maintenance command.
Creating a maintenance mode template before you put the switch in maintenance mode is optional. The objective of maintenance mode for a device is to minimize traffic disruption at the time of removal from the network, as well as during the time of insertion. There are mainly three stages:
-
Graceful removal of the node from network.
-
Performing maintenance on the device.
-
Graceful insertion into the network.
A switch can be put into maintenance mode using default template or a custom template. The default template contains all the ISIS instances, along with shut down l2 . In the custom template, you can configure the required ISIS instances and shutdown l2 option. On entering maintenance mode, all participating protocols are isolated, and L2 ports are shut down. When normal mode is restored, all the protocols and L2 ports are brought back up.
Snapshots are taken automatically while entering and exiting the maintenance mode. You can use the snapshot create snapshot-name snapshot-description command to capture and store snapshots for pre-selected features. Snapshots are useful to compare the state of a switch before it went into maintenance mode and after it came back to normal mode. The snapshot process consists of three parts:
-
Creating a snapshot of the states of a few preselected features on the switch and storing them onthe persistent storage media.
-
Listing the snapshots taken at various time intervals and managing them.
-
Comparing snapshots and showing the summary and details of each feature.
The maximum number of snapshots that may be stored on the switch is 10. You can use the snapshot delete snapshot-name command, to delete a specific snapshot from the device.
You can create multiple templates for the maintenance template or the snapshot template. But only one maintenance template and one snapshot template can be applied to the device at one time.
Snapshot templates can be created to generate specific snapshots. A new snapshot template can be created using the snapshot-templatetemplate-name command. The command snapshot-templatedefault-snapshot-template can be used to specify the default snapshot template in the maintenance mode. The snapshot create[ templatetemplate-name] snapshot-namesnapshot-description command can be used to apply a specific template to the snapshot create feature.
Layer 2 Interface Shutdown
Layer 2 interfaces, such as ports on a switch, are shut down when the system is transitioning into maintenance mode. Layer 2 interfaces are shut down by using the shutdown l2 (maintenance template configuration mode) command in the custom template.
Custom Template
As a network administrator, you can create a template that is applied when the system goes into maintenance mode. This allows you to isolate specific protocols. All instances that need to be isolated must be explicitly specified.
You can create multiple templates with different configurations. However, only a single template is applied to the maintenance mode CLI. Once applied, the template cannot be updated. If the template has to be updated, then you must remove it, make the changes, and then re-apply.
Within a template, protocols belonging to one class are serviced in parallel. The order of priority of the protocols is the same as that of the default template.
To configure this feature, enter the maintenance mode using the system mode maintenance command and enable the feature using the templatetemplate-namecalss command.
For example if the custom template has the following protocols:
Maintenance-template foo
router isis 100
hsrp Et0/1 1
hsrp Et0/1 2
router isis 200
Maintenance-template foo class
router isis 100
hsrp Et0/1 1
hsrp Et0/1 2
router isis 200
In the above example, since isis belongs to CLASS_IGP, router isis 100 & router isis 200 will be serviced in parallel. Once acknowledgements are received for both these protocols belonging to IGP class, FHRP_CLASS clients, hsrp Et0/1 and hsrp Et0/1 2 will be serviced in parallel.
When the template-class feature is configured, the protocols follow an order based on the class they belong to when entering maintenance mode. The protocols follow the opposite order when returning to normal mode.
System Mode Maintenance Counters
GIR has counters to track the following events:
-
Number of times the switch went into maintenance.
-
Ack statistics per client.
-
Nack statistics per client
-
Number of times a particular client did not acknowledge.
-
Number of times switch over happened during GIR. GIR infra will rsync this counter to track multiple switchovers.
-
Number of times the failsafe timer expired.
-
Number of times system got out of maintenance on a timeout expiry.
Enter the show system mode maintenance counters command in privileged EXEC mode, to display the counters that are being tracked by the feature.
Enter the clear system mode maintenance counters command in privileged EXEC mode, to clear the counters supported by the feature.
The client-ack timeout value can be configured using the failsafefailsafe-timeout-value command. Failsafe time is the time that the GIR engine allows a client to transition. Each client sends a notification to the GIR engine about its transition. If it takes more than the failsafe time to transition, it is assumed to have transitioned. The failsafe timer can be configured between 5 - 180 minutes, with a default of 30 minutes.